← Back to Skills Marketplace
xcchenx345

mimo-asr

by xcchenx345 · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ pending
50
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install mimo-asr
Description
小米 MiMo-V2.5-ASR 语音识别技能。通过 Gradio API 调用,将音频转录为文字,支持中英文自动检测和显式指定。无需本地模型,无需 API Key。当用户请求语音转文字、音频转录、提取对话内容时使用。
README (SKILL.md)

MiMo-V2.5-ASR 语音识别 / Speech Recognition (中文/English)

通过小米官方 Gradio 免费 API 调用语音识别模型,无需本地模型、无需 API Key。 Transcribe audio via Xiaomi's free public Gradio API — no local model, no API key needed.

为什么选这个?/ Why This One?

比起安装 4GB 本地模型 + mimo-audio 包,这个版本:

  • ✅ 零依赖(只需要 requests
  • ✅ 零配置(不需要下载任何东西)
  • ✅ 免费无限调用
  • ✅ 支持中文/英文/自动检测
  • ❌ 需要网络(音频传云端处理)

快速开始 / Quick Start

# 安装唯一依赖 / Install the only dependency
pip install requests

# 自动语言检测 / Auto language detection
python scripts/mimo_asr_api.py recording.wav

# 指定中文 / Specify Chinese
python scripts/mimo_asr_api.py speech.mp3 --language zh

# 指定英文 / Specify English
python scripts/mimo_asr_api.py lecture.wav --language en

# 输出到文件 / Output to file
python scripts/mimo_asr_api.py meeting.wav --output transcript.txt

参数 / Parameters

参数 说明 默认值
audio 音频文件路径(wav/mp3/m4a 等) 必填
--language 语言:auto(自动)、zh(中文)、en(英文) auto
--output 输出文本文件路径(可选) 打印到终端

示例 / Examples

# 🪟 Windows
python scripts/mimo_asr_api.py C:\Users\abc15\Desktop\录音.wav --language zh --output output.txt

# 🍎 macOS / 🐧 Linux
python scripts/mimo_asr_api.py ~/Desktop/meeting.mp3 --language auto

API 内部原理 / How It Works

该脚本通过 Gradio API 调用 HuggingFace 上的小米官方空间: The script calls Xiaomi's official HuggingFace space via the Gradio API:

https://xiaomimimo-mimo-v2-5-asr.hf.space

步骤 / Steps:

  1. 上传音频/gradio_api/upload
  2. 启动推理/gradio_api/call/infer(传音频 URL + 语言标签)
  3. 轮询结果 → 获取转录文本

支持的文件格式 / Supported Formats

WAV、MP3、M4A、FLAC、OGG 等主流音频格式。 WAV, MP3, M4A, FLAC, OGG and common audio formats.

注意事项 / Notes

  • 大文件处理时间较长 / Large files take longer
  • 需网络连接 / Internet connection required
  • 不需要 MIMO API Key / No MIMO_API_KEY needed
  • 免费使用 / Free to use
Capability Tags
requires-sensitive-credentials
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install mimo-asr
  3. After installation, invoke the skill by name or use /mimo-asr
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
MiMo-V2.5-ASR Skill v1.0.0 - Initial release of the MiMo-ASR skill for audio-to-text transcription. - Uses Xiaomi's public Gradio API for speech recognition—no local model or API key required. - Supports Chinese, English, and auto language detection. - Compatible with major audio formats (wav, mp3, m4a, flac, ogg, etc.). - Simple setup: only requires the `requests` library and an internet connection. - Free unlimited usage.
Metadata
Slug mimo-asr
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is mimo-asr?

小米 MiMo-V2.5-ASR 语音识别技能。通过 Gradio API 调用,将音频转录为文字,支持中英文自动检测和显式指定。无需本地模型,无需 API Key。当用户请求语音转文字、音频转录、提取对话内容时使用。 It is an AI Agent Skill for Claude Code / OpenClaw, with 50 downloads so far.

How do I install mimo-asr?

Run "/install mimo-asr" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is mimo-asr free?

Yes, mimo-asr is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does mimo-asr support?

mimo-asr is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created mimo-asr?

It is built and maintained by xcchenx345 (@xcchenx345); the current version is v1.0.0.

💬 Comments